Stringent U.S. environmental regulations will in the immediate future require that the level of sulfur in gasoline be reduced by 90% from the current 300 ppm to 30 ppm and those in diesel be reduced by 97% from the current 500 ppm to 15 ppm. Hydrotreating is most common method of removing organic sulfur and nitrogen compounds from petroleum fractions. In hydrotreating, oil and hydrogen are fed to a fixed bed reactor that is packed with a hydrodesulfurization (HDS) catalyst. The HDS operating temperature and pressure typically range from 600-700° F. and from 500 to 2,500 psig (pounds per square inch, gauge), respectively. The more difficult the sulfur removal needed, e.g., the higher the level of sulfur reduction, the more stringent the HDS operating temperatures and pressures become. In this regard, severe hydrotreating of gasoline feedstock to achieve low sulfur levels will saturate a significant portion of the olefins in the gasoline thereby substantially lowering the octane number. To minimize the octane loss, state of the art hydrotreating catalysts can isomerize the paraffins that are generated by olefin saturation. In a similar vein, it is expected that more robust catalysts must be developed and efficient process modifications implemented in order to remove the most refractory sulfur compounds. Most refiners have revamped their existing hydrotreating facilities and/or introduced new hydrotreating techniques in anticipation of these challenges as they comply with the new U.S. guidelines.
In recent years, industry has sought to develop less expensive desulfurization alternatives to hydrotreating. It is known that contacting a petroleum distillate to an oxidant converts sulfur and nitrogen compounds in the distillate into sulfones (or sulfoxides) and organic nitric oxides, respectively. These polar organic oxides can be removed from the distillate by solvent extraction and/or adsorption. More importantly, oxidative desulfurization can easily oxidize and remove thiophenic sulfur compounds, which cannot be readily treated by HDS due to the stereo hindrance effect around the sulfur atom in the molecule. For example, it has been reported that the activity of thiophenic compounds in responding to HDS treatment is in the following sequence: DBT (dibenzothiophene)>4 MDBT (4-methyl dibenzothiophene)>4,6 DMDBT (4,6-dimethyl dibenzothiophene). See, Ind Eng Chem Res, 33, pp 2975-88 (1994). In contrast, it has been reported that the activity of thiophenic compounds in responding to oxidative treatment is just the opposite, namely: 4,6 DMDBT>4 MDBT>DBT. See, Energy Fuels, 14, pp 1232-39 (2000). These observations suggest that oxidative desulfurization can be effective in removing the most difficult residual sulfurs from hydrotreated oils to yield ultra-low sulfur products.
The oxidants currently used in oxidative desulfurization include, for example, peroxy organic acids, catalyzed hydroperoxides and inorganic peroxy acids. Almost all peroxy organic acids are derived by oxidation of organic acids with hydrogen peroxide. For example, EP 1004576 A1 to Druitte discloses a process for producing peracetic acid (PAA) by reacting hydrogen peroxide and acetic acid (AA) in an aqueous reaction medium.
U.S. Pat. No. 6,160,193 to Gore discloses a method for removing sulfur and nitrogen compounds from petroleum distillates, such as light gas oil (diesel) by oxidation with a selective oxidant. The oxidants are divided into three categories: (1) hydrogen peroxide based oxidants, (2) ozone based oxidants, and (3) air or oxygen based oxidants. The preferred oxidant is PAA that is formed by oxidizing glacial AA with 30-50% aqueous hydrogen peroxide. Since the peroxide is in the aqueous phase, a phase transfer agent is required to carry the peroxide from the aqueous phase to the oil phase where it oxidizes the sulfur and nitrogen compounds. The phase transfer, which is the rate-limiting step, significantly slows down the reaction rates. In this case, AA is the phase transfer agent for the oxidation of the sulfur and nitrogen compounds in the light gas oil. A small but not insignificant amount of AA remains in the oil phase in the reactor effluent.
Another disadvantage of using the aqueous oxidant disclosed in U.S. Pat. No. 6,160,193 is that the presence of water in the reactor effluent prevents phase separation of oil from the aqueous acid when the oil feed is vacuum gas oil, atmospheric residual oil, crude oil, or other heavy hydrocarbons. Complicating matters is the fact that the sulfones generated in the oxidation reactor also function as surfactants to inhibit phase separation. The spent AA, which is equivalent to 7 to 10 wt % of the oil feed, cannot be effectively removed from the oil, treated, and recycled without phase separation. The presence of water can also cause a significant portion of the sulfones and organic oxides to precipitate from the reactor effluent. Indeed, solids may form at critical stages in the process thereby causing the valves, pumps, and even the adsorbent bed to malfunction. U.S. Pat. No. 6,160,193 does not appear to recognize the importance of the solid precipitation problem, which certainly occurs when the distillate contains more than 500 ppm sulfur and nitrogen compounds.
The specific solvents used to extract sulfones from the distillate phase in the process disclosed in U.S. Pat. No. 6,160,193 also tend to extract appreciable amounts of oil along with the sulfones and organic nitrogen oxides. The prior art has disclosed many solvents for the sulfones extraction, including dimethyl sulfoxide (DMSO), formic acid, nitromethane, dimethyl formamide (DMF) and trimethyl phosphate. See, for example, U.S. Pat. No. 6,160,193 to Gore, U.S. Pat. No. 6,274,785 to Gore, U.S. Pat. No. 6,402,940 to Rappas, U.S. Pat. No. 6,406,616 to Rappas et al., and EP 0565324 A1 to Aida. However, none of these solvents has proven to be cost effective in removing sulfones from the oil.
U.S. Pat. No. 6,402,940 to Rappas describes a process for desulfurizing fuels such as diesel oil to achieve a sulfur level of 2 to 15 ppm. The oxidant is hydrogen peroxide in a formic acid solution with no more than 25 wt % water. Since hydrogen peroxide is in an aqueous phase, the formic acid functions as the phase transfer agent that transfers the hydrogen peroxide to the oil phase. Given that formic acid is a more efficient phase transfer agent than acetic acid, the oxidation reaction rate is faster under formic acid. Nevertheless, phase transfer remains the rate-limiting step. A major drawbacks of the process is the spent acid recovery system. As described in the patent, the spent acid, which contains formic acid, water, sulfones, and trace amounts of diesel, is first fed to a flash distillation vessel to strip out the formic acid and water. The formic acid and water are then fed to an azeotropic distillation column. In this process, water is derived from oxidation reactions and from the aqueous hydrogen peroxide feed. Water must be removed from the spent formic acid stream in order to maintain the water balance in the process. It is known that formic acid and water form an azeotrope containing 77.5 wt % formic acid and 22.5 wt % water. However, according to the disclosed process, feed to the azeotropic distillation column contains more than 77.5 wt % formic acid. Consequently, the column could produce essentially pure formic acid in the overhead stream and about 77.5 wt % formic acid (but not pure water) in the bottom stream. In light of this, it would be impossible to remove water from the spent formic acid and it appears that the disclosed process is inoperable.
The presence of water in the reactor effluent also causes a significant portion of the sulfones and organic oxides to precipitate from the liquid phases and thereby disrupt the process. As mentioned earlier, water in the system also renders the process unsuitable for desulfurizing heavy hydrocarbons, such as vacuum gas oil, atmospheric resid, and crude oil, due to the difficulties in phase separation between oil and the aqueous acid.